Cu 2 ZnSnS 4 (CZTS) is a promising material for thin film solar cells based on sustainable resources. This paper explores some consequences of the chemical instability between CZTS and the standard Mo "back contact" layer used in the solar cell. Chemical passivation of the back contact interface using titanium nitride (TiN) diffusion barriers, combined with variations in the CZTS annealing process, enables us to isolate the effects of back contact chemistry on the electrical properties of the CZTS layer that result from the synthesis, as determined by measurements on completed solar cells. It is found that instability in the back contact is responsible for large current losses in the finished solar cell, which can be distinguished from other losses that arise from instabilities in the surface of the CZTS layer during annealing. The TiN-passivated back contact is an effective barrier to sulfur atoms and therefore prevents reactions between CZTS and Mo. However, it also results in a high series resistance and thus a reduced fill factor in the solar cell. The need for high chalcogen pressure during CZTS annealing can be linked to suppression of the back contact reactions and could potentially be avoided if better inert back contacts were to be developed.

No abstract

Defect energy levels in hydrogen-implanted and electron-irradiated n -type 4H silicon carbideThe annealing behavior of irradiation-induced defects in 4H-SiC epitaxial layers grown by chemical-vapor deposition has been systematically studied by means of deep level transient spectroscopy ͑DLTS͒. The nitrogen-doped epitaxial layers have been irradiated with 15-MeV electrons at room temperature and an isochronal annealing series from 100 to 2000°C has been performed. The DLTS measurements, which have been carried out in the temperature range from 120 to 630 K after each annealing step, revealed the presence of six electron traps located in the energy range of 0.45-1.6 eV below the conduction-band edge ͑E c ͒. The most prominent and stable ones occur at E c − 0.70 eV ͑labeled Z 1/2 ͒ and E c − 1.60 eV͑EH 6/7 ͒. After exhibiting a multistage annealing process over a wide temperature range, presumably caused by reactions with migrating defects, a significant fraction of both Z 1/2 and EH 6/7 ͑25%͒ still persists at 2000°C and activation energies for dissociation in excess of 8 and ϳ7.5 eV are estimated for Z 1/2 and EH 6/7 , respectively. On the basis of these results, the identity of Z 1/2 and EH 6/7 is discussed and related to previous assignments in the literature.

No abstract

Comparative Hall effect investigations are conducted on N- and P-implanted as well as on (N+P)-coimplanted 4H–SiC epilayers. Box profiles with three different mean concentrations ranging from 2.5×1018 to 3×1020 cm−3 to a depth of 0.8 μm are implanted at 500 °C into the (0001)-face of the initially p-type (Al-doped) epilayers. Postimplantation anneals at 1700 °C for 30 min are conducted to electrically activate the implanted N+ and P+ ions. Our systematic Hall effect investigations demonstrate that there is a critical donor concentration of (2–5)×1019 cm−3. Below this value, N- and P-donors result in comparable sheet resistances. The critical concentration represents an upper limit for electrically active N donors, while P donors can be activated at concentrations above 1020 cm−3. This high concentration of electrically active P donors is responsible for the observed low sheet resistance of 35 Ω/□, which is about one order of magnitude lower than the minimum sheet resistance achieved by N implantation.